Grinding & Classification Circuits

Grinding & Classification Circuits 2017-04-04T06:57:16+00:00
  • To participate in the 911Metallurgist Forums, be sure to JOINLOGIN
  • Use Add New Topic to ask a New Question/Discussion about Grinding.
  • OR Select a Topic that Interests you.
  • Use Add Reply = to Reply/Participate in a Topic/Discussion (most frequent).
    Using Add Reply allows you to Attach Images or PDF files and provide a more complete input.
  • Use Add Comment = to comment on someone else’s Reply in an already active Topic/Discussion.

HPGR vs SAG (64 replies and 1 comment)

Bill Rico
1 year ago
Bill Rico 1 year ago

Hello: I would like to start a new discussion. When HPGR should be used instead of SAGs in a grinding circuit. What are the technical advantages, sensitivity to feed distribution, hardness of ore, efficiency etc? What are the trade-offs between SAG vs HPGR grinding technologies?

Carmen Ibanz
1 year ago
Carmen Ibanz 1 year ago

HPGRs are suitable for comminution circuits that deal with ores qualified as "competent". I like to leave the term hardness out of this analysis so I recommend not using the Bond Mill work index as an indicator to decide whether to use HPGRs in a circuit or not.

Everyone in the industry is trying to have the simplest circuit possible that will deliver the lowest CAPEX and OPEX possible while allowing to reach design capacity but in case of competent ores there is no alternative other than adding "crushing power". With technology available today you will be forced to have about 3 stages of crushing before feeding a ball mill if your ore is a competent ore.
HPGR is a machine that plays a role in tertiary crushing, receiving top sizes of about 45 mm and delivering product sizes of around 4 mm (this is achieved by using a classification circuit together with the HPGRs). I have seen HPGRs in circuits dealing with throughput of about 1,500 t/h of fresh feed, so you can safely assume that HPGR can be used in a wide variety of tonnages up to the limits I mention. In one place the circuit I have seen working has 3-4 HPGRs to process about 4,800 t/h.
May be an easiest way to show when HPGRs are needed in a circuit is to look at those projects where a SAG mill circuit didn't reach design capacity. If I remember well ALL of the SAG circuits dealing with competent ores have been forced to add either pebble crushers or secondary crushers to reach design capacity. This situation has transformed some SAG mills into an “unusual” hybrid mill which looks like a SAG but that is fed with a particle size close to the typical critical size that forms inside a SAG mill, i.e particle sizes in the range of 45 mm. The critical size generated in a SAG mill when dealing with competent ores is such that these pebbles are very hard to crush and impossible to grind down in a reasonable amount of time in a regular tumbling mill. The 45 mm competent ore produced in a crusher has internal fractures that make it suitable to feed a SAG mill where these rocks are easily reduced to final SAG product.

David
1 year ago

HPGR vs SAG ... great discussion.

Jean Rasczak
1 year ago
Jean Rasczak 1 year ago

In discussion with about this topic some time ago I asked him a similar question. He suggested the following high level guidance when thinking about the selection of equipment to treat particular ore strengths.

"Rotating mills, conventional crushers and HPGR's can be used depending on the rock strength. If the multiple of the drop test parameters ‘a’ and ‘b’ (AxB) is

• Above 200 use a scrubber
• Above 60 use AG milling
• Above 40 use SAG milling
• Above 25 use HPGR
AxB very rarely goes below 25. For AG and HPRC analysis the SMC test and DWi is very important." Mike may care to clarify further?

An advantage of HPGR's is the ability to get high breakage energy (kWhr/t) into the rock in one pass resulting in lower circulating loads.

Carmen Ibanz
1 year ago
Carmen Ibanz 1 year ago

I recommend thinking on HPGR for AxB below 40. At AxB values between 37 and 40 there may still be a trade-off between conventional SAG and HPGR circuits. Under 37 I would be very careful to have a SAG circuit in my design. In this case a secondary crusher or pebble crusher (together with big SAG grate openings) is almost a must have.

Marshal Meru
1 year ago
Marshal Meru 1 year ago

There has been a suggestion that HPGR will replace SAG milling in the future. 

Anyone have any experience in that regard, i.e. estimation of CAPEX and OPEX versus SAG’s?
Presently, it appears that HPGR is replacing cone crushers in a SAG circuit to crush hard SAG pebble.

Carmen Ibanz
1 year ago
Carmen Ibanz 1 year ago

HPGR may not be the right machine for less competent ore, especially when the ore has clays. So far it looks like that there are certain ore types that require too much energy to have them ground down in a conventional SAG-Ball Mill circuit. This is the case where HPGR circuits come handy. When the ore is competent there are less fractures available so the ore gets its size reduction mainly through abrasion. You will understand that trying to abrade a 6 inches rock down to 2.5 inches may take lots of time and hence the rock just spins around the SAG mill. If you want to reach design capacity then you need more SAG mills which just take you down the path of energy waste.

For less competent ore the conventional circuit is still the best choice due to lower CAPEX.
HPGR circuits needs that more real state is made available as an ore classifying system needs to be in place. This system requires the construction of conveyor belt systems which enlarge the size of a circuit.

Marshal Dienes
1 year ago
Marshal Dienes 1 year ago

I think, the SAG/FAG cannot simply be compared with HPGR, taking account of their very different grinding process requirements and results, even the process is dry. As an example, if you expect to grind an ore from 200 mm size to 1 mm size, the use of the HPGR instead SAG/FAG requires the addition of 1…2 crushing phases, consequently the FAG/SAG cannot be replaced by HPGR only. If we add the screening, handling and feeding equipment CAPEX and OPEX, the comparison work will be more complex. The complexity degree of this comparison will be higher if the process water is available, due to the significant benefits of the wet screening vs. dry screening. Another aspect, with a major impact on the screening efficiency, is the different particle forms resulted from the SAG and HPRG processes. In conclusion, the optimal grinding circuit and its equipment requires a complex study of the technical and financial aspects. Taking into consideration the different recommended utilization areas of SAG and HPGR, their limited comparison is not possible to be correctly developed.

Carl Jenkins
1 year ago
Carl Jenkins 1 year ago

Now I finished the calculation of flowsheet with pre-crushing, SAG mill and Ball mills.

The result is follows:

1. Pre-crushing- three crusher MP-2500(Metso) for crushing from 300mm to 50mm. Two working and once in reserve.
2. Primary grinding- once SAG mill with size 42ft x 24ft, power motor 28MW. F80-50mm, p80- 2mm.
3. Secondary grinding- two Ball mills with size 31ft x 45ft, power motor 28MW (each). F80- 2mm, P80 - 250microns.
4. Thirty grinding- once Ball mill (analogically). F80- 350microns, P80- 150microns.
5. Capacity of flowsheet - 10 000t/h.
6. Bond index for ball mill- 12.7 KW/t.
7. Power consumption of flowsheet - 10KW/t.

This calculations are made by modeling program "AKIAN".
The interestingly that "AKIAN" showed is necessary thirty grinding.

My question - how much biggest HPGR mills are necessary for replace this SAG mill?

Zander Barcalow
1 year ago
Zander Barcalow 1 year ago

You can think to the Weir Minerals/KHD HPGR 16-170/180..

Carmen Ibanz
1 year ago
Carmen Ibanz 1 year ago

So far we have clarified that HPGR doesn't replace s SAG mill and hence these two machines can't be compared. We need to compare the HPGR circuits against the SAG-crush-Ball mill and against the Crush--Ball mill circuits 

Metso has the largest HPGR in the world today. I suggest to Contact Metso.

There are numerous papers related to this matter.

Ace Levy
1 year ago
Ace Levy 1 year ago

A always with equipment differences "there are horses for courses" - an analogy I would like to propose is that HPGR is like a race horse, not too good in variable circumstances in the same race, no hills, no uneven ground and a fixed handicap as the main challenge to performance. All this being correct, it is the fastest horse available option to get from point A to point B.

Now a SAG, could be seen as, a pack horse, a rodeo quarter horse, show jumper and all those other equine functions that require flexibility. One clear example is the ability to change it into a large ball mill (seems quite common these days) with increased charge and higher dependence of a P80 particle reduction.
Drawing a even longer bow, you tend to shoot a race horse, should it break a leg....and spend serious money to replace it....

Apologies for the simplicity, but it may have some merit as a starting point.

Sugar Watkins
1 year ago
Sugar Watkins 1 year ago

An interesting question. Published data shows HPGR capex is typically 15% to 30% higher than a comparable SABC circuit for the hardest ores. HPGR though provides opex reductions. I am curious to see if the opex reductions have actually panned out? Any experience out there on this?

Carmen Ibanz
1 year ago
Carmen Ibanz 1 year ago

Cerro Verde reports lower energy consumption when compared to SABC circuits

Cerro Verde actual data shows that the HPGR circuit energy consumption was 3 kWh/t less than the SAG/Ball mill circuit that was considered in the trade off studies.
The initial trade off studies were expecting to have 4.2 kWH/t of less energy consumption.

Carl Jenkins
1 year ago
Carl Jenkins 1 year ago

These time the favorite horse is SAG. 

HPGR may be after 5...10 years. Why? SAG mill drinks water more them HPGR. Water consumption of SAG nearly 1.0 m3/t (with thickener), HPGR have less 0.1m3/h (without thickener). After 5...10 years the capacity plant will be 70...100 Mt/y may be more. Where to look 100 million tons water?
May be after 5 years will be new favorite horse. For example - microwave grinder (Pegasus).

In this discussion I offer to compare three actual flowsheet or three horses:

1. Pre-crushing-HPGR - ball mill (for example Baddington).
2. SAG- pebble crashing- ball mill (Toromocha, Peru).
3. Pre-crushing- SAG- ball mill. (Meadowbank, Canada).

Carmen Ibanz
1 year ago
Carmen Ibanz 1 year ago

I would recommend to see if the horse would do the work he is expected to do. You may choose a cheaper horse that may look strong but if he won't walk at the pace you want THEN you may not arrive to the cash flow town you expect. 

The decision needs to include a risk analysis. My suggestion is to look at the rock "competency" or look at the axb. If axb is low I would recommend to NOT believe in the SAG design output and add some extra percentage of power needed, let’s say 20% or 30% more. The reason for this is that SAG modeling tools available have the tendency to make the wrong predictions when SAG circuits deal with competent ore.

Carl Jenkins
1 year ago
Carl Jenkins 1 year ago

Please give me actual data without capacity for anyone SAG mill. In own hour I tell you the capacity of this mill.

Jean Rasczak
1 year ago
Jean Rasczak 1 year ago

I always enjoy reading your posts as all of you get the center of the matter quickly.

However, things are changing with SAG throughputs regularly achieving higher than rated flows and better P80 expectations. At the risk of not trying to promote our work, as these forums are not for this at all.
Our work these days is actually blending ball sizes and grades to optimize the process. This is not usual as most ball makers would not be able to blend say a 5" ball with a 3" ball in a high impact SAG environment without killing the 3" media either partly or totally. This now can be done without media carnage - plus the media retains its spherical shape throughout its deployment.
Take a look at this scenario - you can actually have your cake and eat it too.
I believe I am keeping within the scope of the question as it does give a SAG scenario that is not possible with HPGR that has no media to play with. We calculate impact energy in relation to ball size ensuring that targeted throughputs are met - then using the grinding surface adjustability of blending we can target the grind closely as well.
The question that I would like to put forward to this forum is; As SAG can have this little spoken about flexibility, how can HPGR ever hope to match this?

Carmen Ibanz
1 year ago
Carmen Ibanz 1 year ago

Let’s assume AxB of 32, Bond mill work index of 20, Crusher work index of 10. Is this enough data?

Carl Jenkins
1 year ago
Carl Jenkins 1 year ago

Size (in meters) of sag mill, size & load of balls, F80 & T80, speed of SAG

Carmen Ibanz
1 year ago
Carmen Ibanz 1 year ago

SAG Size: 12.8 x 7.6

size of balls in SAG: 6 inch
size of balls in Ball Mill: 3 inch (50%) 2 in (50%)
load of balls in SAG: 14%
load of balls in Ball mills: 32%
f80 to SAG: 152 mm (6 inches)
SAG P80: 2 inch
Circuit p80: 274 microns
SAG speed : 10 RPM

Carl Jenkins
1 year ago
Carl Jenkins 1 year ago

Tell methe size & number & speed ball mills, "SAG p80-2 in" or 2mm?

Carmen Ibanz
1 year ago
Carmen Ibanz 1 year ago

This scenario for the SAG mill is taken from a design criteria. P80 for the SAG is 2 inch (this p80 considers SAG discharge only). If we consider the SAG screen and pebble crusher then SAG CIRCUIT p80 = 7mm.

Ball Mills: 2
Ball Mill speed: 75% of critical speed.
Ball mill size: 27 x 45 ft

Marshal Meru
1 year ago
Marshal Meru 1 year ago

This one is not a design that I have encountered before which is interesting as competent 150mm SAG media can only be made by a very, very limited list of suppliers. We have theoretically designed 200mm media as an exercise targeting almost 60 as an average volumetric hardness. We may look at the modeling for a size blend which may also include a percentage of 140mm media as well for that finer P80. !2.8m SAG is also an odd one. First thought, apart from the above, is a P80 of 50mm (2") when say 15mm/20mm could be achievable even at that bond work index of 20kWt/h. as a direct mill grate discharge prior pebble crushing.

Frankly, I would really like to have a close look at this ourselves if it ever got passed the design stage. The Ball mills are a great size as well and probably very capable of better than the 274um noted as the final grind or BM discharge. A blended media size with compatible ball grades would also be interesting to run through our modeling. Totally, very interesting sets of equipment to fine tune.

Carl Jenkins
1 year ago
Carl Jenkins 1 year ago

3600 t/h.

Carmen Ibanz
1 year ago
Carmen Ibanz 1 year ago

I agree with your number. 

I have seen a design for a similar ore who has 4 HPGRs (5,600 kw each) and 4 cone crushers (933kw each) in the secondary crusher area. It has a 60x113 primary crusher and 3 ball mills (17,000 kw each). This circuit process between 110-120 ktpd

I'm not looking to get advice on designing a new circuit but to only add facts (numbers from actual operations) to this discussion to try help Tony with his question
I appreciate your passion and help with the exchange of information.
Knowledge is the only thing we will leave in this world long after we are gone.

Marshal Dienes
1 year ago
Marshal Dienes 1 year ago

Sometimes a radical proposal is the best analogy to clarify the point you are making.

As a media supplier, I am happy to concede applications where HPGR is indeed the front runner in equipment selection. HPGR will be the better performer when you have a very consistent ore body as that machine can be specifically set up to maximize the greater ore quantities. It is when the ore body feed to processing has a wide range of ore variations that fast processing adjustment and flexibility is demanded. Then SAG takes the lead.

Ace Levy
1 year ago
Ace Levy 1 year ago

SAG-based circuits incur lower capital costs while HPGR-based circuits incur lower operating costs. There is a need for Trade-off study.

The benefits of the HPGR:

• Higher energy-efficiency than SAG milling
• Finer product than cone crushing
• Similar transfer size to SAG milling
• Reduced work index of product
• Improved minerals liberation
• Reduced over-grinding and sliming
• Smaller footprint (m²/kW)
• Improved delivery times
• Shorter installation times
• More rapid plant ramp-up
• Easier plant debottlenecking
Disadvantages:
• Plant complexity, larger overall footprint:
• Closed circuit secondary and HPGR
• Crushers & screens
• Conveyors, bins & feeders
• Dust control
• Tramp metal management
Capital costs:
Capex differential = ƒ(ore competency-1)
Capex differential = ƒ(plant capacity-1)
100,000 tpd – 10%
20,000 tpd – 30%Show less

Carl Jenkins
1 year ago
Carl Jenkins 1 year ago

Thanks on example of SAG mill flowsheet. The design criteria of this flowsheet i investigated (with "AKIAN") and head follow result:

1. Capacity of flowsheet - 3600t/h. This is few for SAG mill with size 12.8 meters of diameter and 7.4 meters of light (even with 20 Bond Index).
2. The work power consumption of SAG is 20,24 MW. Nearly 80% (16.3 MW) consumed to crushing large rocks from 350mm ( F80=152mm) to 56mm. Specific power consumption will be 4.3 kW/t. May be this work make in secondary crushers ( 0.7kW/t)?
3. The work power consumption of ball mills is 21.2 MW (each) and capacity of each mill is 1800 t/h. This is too few.

" AKIAN" calculated the other flowsheet without pebble crushing with follow criteria:
1. Feed size for SAG mill - 60mm.
2. Transfer size for SAG mill- 2mm.
3. Capacity flowsheet- 4800t/h.

Mark, i understand that you want blending difference ball size for increasing capacity of mill and reducing P80? We investigated at this and head follow result. You are right about killing small balls by large balls. Ball of 4' are will killing the 1.5' ball. Therefore blend 4' & 1.5' is impossible. If use the 2'ball the lifetime of it will be short. If use only 4' so after 20...25 days the balls size curve is will be consist 80% less 3.2' (with permanent adding necessary 4' balls) and 15...20% less 1.5'. This curve will be permanently. If blending 4' with 2.5' the grade of 1.5' will be more 20%. Is it need?
I think for SAG mill the load ball size will be one size.
For reducing P80 this work must do Ball mill (secondary grinding) with small balls. We are load only 2.5' and this ball killed 1' ball. For more reducing P80 necessary use thirty grinding with Ball mill or Vertimill with load of 1.0'...1.5'ball etc.

Dizzy Flores
1 year ago
Dizzy Flores 1 year ago

I am not using computer programs, I leave that to our technical team when we need to make ball selections with high accuracy.

However, some good up to date news for you! We have the ability to make even 5" with 3" in a blend that works in harmony - without breakage concerns.
As a SAG guy you will appreciate this as a new set of options.
This plan was first started in ball mills we tune, in an actual trial account, this was a start-up seasoned charge in a single stage mill where we had several sizes - the top being a 133mm and the smallest being 80mm. Commissioning was quicker and the results were very good.

I do think blending ball sizes with compatible media grades is the future and a flexibility that HPGR cannot match. At this time we are carefully planning a shift in the final grind by using a 2.8: ration of 115mm: 90mm. This should move 76% of passing 200mesh to over 80%. I would welcome your comments on this as well.

Keeping on topic regarding what SAG can do when the final grind tuning is being targeted which HPGR cannot.

Carl Jenkins
1 year ago
Carl Jenkins 1 year ago

My information about blending balls only for information for you.

Zander Barcalow
1 year ago
Zander Barcalow 1 year ago

My comments from metallurgical point of view:-

1. For the time being let us forget about CAPEX OPEX , Capacity, Efficiency. of SAG Vs HPGR. Let us talk something different. Particle size shape after grinding.

  • i)SAG: It produces lot of slime which will effect grade and recovery in down line operation of flotation, leaching, etc. We say particle size in terms of passing # in %. Say 75 micron. But in this 20 micron and sub-micron particles are more, in SAG mill. We need closely sized particles well liberated, and economical.
  • ii)ii) HPGR: For soft ores it can reduce up to 75 micron, and hard ores 1mm. Today many down line process is successful at primary grinding level of 75 micron to 1mm. This can solve many problems of handling and metallurgical problems of processing ores. Examples. In Iron ore Hematite if you grind more recovery is less, Blaine if it exceeds 2000 pellet costing and quality goes high. Many pellet industries are facing burning problems with this Blain number. No one has solved the problems till to day.
    iii) Operating parameters are less in HPGR. Consumables are less. Supervision and controls are less. Simple and easy to understand. Designing HPGR is very simple.
    iv) Let us have few applications for discussion. Just theory is not correct. Many theories failed during operation. Even equipment's were discarded later.

In Lead and Zinc grinding Lead is soft. SAG will develop lot of sub-micron particles which are lost in tailing. Who will give answer for this?

Carmen Ibanz
1 year ago
Carmen Ibanz 1 year ago

I think that the real evaluation is not only CAPEX and OPEX or Metallurgical Performance. 

You will be amazed to see people making decisions based on CAPEX only and based on what they are used to do. If they have been operating with a SAG circuit they will have the tendency to continue doing this.
The real evaluation comes from developing a trade-off where the following needs to be considered
CAPEX and OPEX
Revenues (coming from Metallurgical performance as Kshirasagara points out)
Risks of project construction (for example, is the new site far from roads infrastructure? this puts a limitation on machine sizes sometimes)
Availability of financing
Investors expectations in regards to the time they will have their capital back and start making earnings.
Start-up Risks (learning curve of the operators)
Social Risks (dust generation for example)

Sugar Watkins
1 year ago
Sugar Watkins 1 year ago

I studied all comments related SAG vs. HPGR. In addition, I remarked the change of flowsheet production capacity, from 10000 t/h to 3600…4000 t/h. I strongly recommend you, in accordance with Juan’s opinion, to select the following comminution steps: 1) Crushing, from 300 to 40 mm; 2) HPGRs from F80 40 mm to P80 2 mm in closed circuit with dry screening 2 mm; 3) The under screening -2 mm will be submitted to the wet screening 0.25 mm opening; the size fractions +0.25 mm will be sent to the secondary BM and the under screening to tertiary grinding; 4) Secondary BM grinding, F80 2 mm, P80 0.25 mm; 5) Tertiary BM grinding, F80 0.25 mm, P80 0.150 mm.

Taking account of your lower production capacity, you can think to replace the two BMG phases by one VM grinding step only, able to ensure a very efficient grinding from F80 2 mm to P80 0.150 mm under very, very efficient technical and financial conditions. You can look at METSO VM 3000 or new METSO VM 5000. You can develop the VM grinding tests with METSO (USA, PA) on 60 kg ore sample and the cost of the test is about US$6000. Based on the test result you can quantify the accurate number of VMs required by your ore and production capacity. The VM is a very profitable grinding equipment, under incidence of the OPEX and even the CAPEX. Based on my rough estimate, you need 14…16 VMs for your flowsheet of 3600 tph. If you need further information concerning the VMs and their high grinding efficiency, please contact me. It will my pleasure to introduce you to METSO people, USA, PA, in order to develop the VM grinding test.

Jean Rasczak
1 year ago
Jean Rasczak 1 year ago

I did very quick calculation about your circuit. Am I missing anything? I calculated your power demand from available data and the power that can be supplied from your equipment list. 

(Assuming MP 2500 is 2500 HP or 1864 kW and tertiary ball mill equipment power is 20% more than its power demand of 35 MW).

Unit Eq Power Total Power Supplied
kW kW
MP-2500(Metso) 2 1,864 3,728
SAG mill 1 28,000 28,000
Ball mills 2 28,000 56,000
Thirty grinding- Ball mill 1 42,973 42,973
Total (kW) 130,701
Specific Energy (kwh/t) 13.07

Your specific energy need is 10 kwh/t (that is 75% of my calculation) is that correct? 

Zander Barcalow
1 year ago
Zander Barcalow 1 year ago

Certainly, there will be some saving on power consumption when HPGR is used as against a SAG Mill. However, there are other factors that do affect the decision, as well. For example, abrasiveness of the ore.

We have just completed DFS for a large Zinc Mine with beneficiation plant. Carried out a trade-off between SAG Mill and HPGR. Simply because of high abrasiveness of the ore (Index = 65) the saving on power got set-off by higher consumption of wear components (Crusher Liners & Screen panels) and requirement of additional O&M staff.

Ace Levy
1 year ago
Ace Levy 1 year ago

Make decision on commercial recovery and grade not on equipment based. Finally it is the operator and owner who is benefited and not equipment vendor. After SAG if metallurgy fails then what to do. Who will answer, vendor or operator? Today many industries are facing serious problems in this subject. Test both particles produced form SAG and HPGR and do cost benefit analysis and then take decision. It is the final Cost of concentrate produced, quality, and quantity that will give you confidence in buying equipment. Please don t forget metallurgy part.

Carl Jenkins
1 year ago
Carl Jenkins 1 year ago

Secondary crushers no exist in total specific power. sag-28mw (install), 24mw (working). sec.mills 28+28=56(install) 25+25(work) tertiary. 28(i) 26(w). tot=24+50+26=100mw 100/10 =10kw/t

Dizzy Flores
1 year ago
Dizzy Flores 1 year ago

As a balance, what would suggest for a SAG circuit? If there is more than one option from you it would sound less like an advertisement for Metso.

Please do not take offence to these remarks, as it is not meant to be so, the circuit you have mapped seems quite viable BUT an alternate SAG circuit proposal (as you see it) will make it less of an "only" alternative. VM's are also interesting to look at, as they do seem very efficient.
A single stage BM we have been fine tuning has F80: 75mm and P80+: passing 200mesh and does it easily. Much better than 2mm to 150um in a single step.

Marshal Meru
1 year ago
Marshal Meru 1 year ago

You can select your flowsheet based on BM use but, as I mentioned, it is recommendable to replace the SAG equipment (F80 50 mm; P80 2 mm) by HPGRs keeping the same feed and product sizes 50 mm and 2 mm). It is your choice but, taking account of your new throughput (4000 t/h), you can look at another flowsheet variant, significantly more efficient (lower OPEX) and characterized by reduced CAPEX than comminution scenario SAG + 2 BM stages. I suggest you to consider the following grinding circuit:

1. Primary HPRG F80 50 mm, P80 3 mm, in closed circuit with dry screening 3 mm opening (3 mm instead 2 mm in order to increase the dry screening efficiency);
2. Secondary grinding, using the Metso Vertimill 3000 (F80 3 mm and P80 250 µm, the last being your required final product size).

Under these conditions, you will have 2 grinding steps only, the bothering characterized by lower OPEX and CAPEX than your scenario.

My flowsheet proposal is based on the following criteria:

1. The Metso VMT is characterized by lower power consumption than BM (up to 35 % less power) and lower grinding media consumption than BM (up to 53% less media consumption);
2. The operation of your secondary and tertiary BMs is not situated on the optimal size range of the product size (BM optimal product size 15000…900 µM) in comparison with Metso VTM optimal operating range, of 1000…40 µm; You need the final size 250 µm;
3. The CAPEX and OPEX of the fine grinding circuit will be considerably reduced (screens, feeders, pipes etc.).

In order to develop this flowsheet and accurately assess the CAPEX and OPEX, you need to proceed to HPRG laboratory and, in addition, the semi-pilot tests. The VTM tests will be developed with METSO, on the ground sample resulted from HPRG semi-pilot test (3 mm size) in order to check the VTM response to the size distribution obtained from HPRG process (tests). If you need further information please do not hesitate to contact me.

Carl Jenkins
1 year ago
Carl Jenkins 1 year ago

To replace the Ball Mill by HP, yes SAG, no.

Carmen Ibanz
1 year ago
Carmen Ibanz 1 year ago

I case that you don't have it, I suggest to take a look to this paper.

What can go wrong in comminution circuit design.

by C Bailey, , G Lane, , S Morrell and P Staples

http://www.ceecthefuture.org/wp-content/uploads/2012/11/Baileygowrong-11.pdf

and this other one

http://www.sgs.ca/~/media/Global/Documents/Technical%20Documents/SGS%20Technical%20Papers/SGS%20MIN%20TP2011%2008%20SAG%20Mill%20Testing%20for%20Ore%20Grindability.pdf

Carl Jenkins
1 year ago
Carl Jenkins 1 year ago

Thank you for the paper about vertmill. of course VM more affect them BM on size from 2...3 mm to less. But now the capacity of it few (max 350 t/h). Now this is problem.

Marshal Dienes
1 year ago
Marshal Dienes 1 year ago

I agree with you but please take into consideration the following aspects:

1. The VMs replace 4 BM units (2 secondary and 2 tertiary) based on your assessment; The BM CAPEX are higher than 10...14 VM (about US$10 M each); The accurate VM number can be quantified based on the stirred mill test results; In addition, you will reduce the CAPEX and OPEX required by BM size classification Equipment;

2. The VM OPEX are significantly lower than the BM OPEX; Your BMs, secondary and tertiary, based on their required feed and product sizes, are not operated in their optimal operating ranges.

I do not try to convict you to use the VMs (same situation HPGR instead SAG), but your SAG and BM grinding circuit is an old grinding concept. The final flowsheet, as I mentioned, is your choice. As an example, I would agree to use the SAG if the size reduction ratio, in one step would be 100 or higher. In addition, I kept in my mind your remark concerning the water consumption, but the minimal information package, that you provided, does not allow to select the optimal grinding circuit.

Carmen Ibanz
1 year ago
Carmen Ibanz 1 year ago

A recent evaluation I saw shows that for large tonnages the VM Capex may be as twice as much as capex for BM because more VTMS are needed to replace 1, 2 or 3 BM, whatever is the case

The interesting thing is that savings in maintenance manpower, steel balls and energy largely pay to have more VTMs.
The payback of the extra capex is about 40 months, which is paid by the operational savings.

OPEX for Vertimil may reach values that are 18% lower than for BM circuits
Now some equipment manufacturers will help to pay the initial capex to receive a payment every month for 4 years or so.
I would rather have more machines to operate to have lower OPEX for the mine of the life

Plus the VTMs are getting bigger. Metso is now trying to come up with the VTM 6000. The VTM 4500 is already out there.

Ace Levy
1 year ago
Ace Levy 1 year ago

I agree, but our assumptions have a general character only. The sensitivity analysis of the grinding circuit (NPV and IRR) can provide the accurate response to the question «VM or BM?» In principle, in the case of the size reduction from 2 mm to 0.25 mm (2 stages BM), one stage VM seems to be the better variant. The 4500 Metso VM can successfully use.

Carl Jenkins
1 year ago
Carl Jenkins 1 year ago

SAG, HPRG, BM, VM and other industrial grinders are one generation - mechanism of energy passing are same. Primary energy from electricity net pass to motor, it pass energy to gearbox, then to drum, from it to ball, and only these acts energy (balls kinetic energy) passed to address - to rock or particle ore. If the primary energy is 100%, gearbox consumption will be 10%, consumption drum rotating 35%, to lifting balls, ore and water 55%, consumption to heating of drum, balls, ore, water, grinder "music"& "dance" 15%, to grinding rock or particle ore...no? This is very "effectively", ok? 

This generation is use kinetic energy by contact. The main problem is in comminution, other problem is residence time.
At present time, i think, is necessary determine the concepts for new generation?
In my opinion the new grinder mast be follow:
1. The energy passing to ore particle, will be without contact. For example by microwave, ultrasonic, pneumatic energy etc.
2. The manufacturing material of grinder, will be non-metallic (polymer, ceramic or carbon).
3. The residence time, will be less 1 minute.
4. Needless of water.
5. The grinding and separating of ore particles, will be in one time with one stage (one machine).
6. Installation of grinder will be self-mobile (may be fly).
7. Energy feed will be autonomy (plasma generator with buttery).

Jean Rasczak
1 year ago
Jean Rasczak 1 year ago

I appreciate your efforts and understand them. Your project is a big investment. As Juan and others mentioned, the right way of your work is the preliminary technical and financial study of the both variants of the comminution circuit as follows:

1. Variant A - Crushing (from 300 to 50 mm); SAG (from 50 mm to 3 mm); VM grinding (from 2 to 0.25 mm);

2. Variant B - Crushing (from 300 to 40 mm); HPRG (from 40 mm to 3 mm); VM grinding (from 3 to 0.25 mm).

If the Variant B will be more profitable, I suggest you to develop the HPRG laboratory and pilot tests with the HPGR supplier (i.e. Thyssen Krupp Polysius), on 1.5…2 t crushed ore 300 mm size. The objective of the tests is the complete qualification of the response of your ore to the HPRG process. FYI information, the cost of the VM and HPRG tests are not expensive (VM test on 60…70 kg sample, of 3 mm size – about US$6000; HPRG test (laboratory and pilot, in closed circuit, about US$25000). For additional data and test development, you can contact your regional representative of the suppliers of HPGRs and VMs (Metso). Please carefully study and select the optimal screening concepts and equipment, under incidence of CAPEX and OPEX, in addition to technical criteria.

Dizzy Flores
1 year ago
Dizzy Flores 1 year ago

We do have an exceptionally good group of calculators in this forum!

I have always wanted someone with this ability (I do not have it) to look at a hypothetical case, I say this as we are travelling down this road now and it would be wonderful if anyone could be able to calculate the possible efficiency outcome.
SAG/BM circuit.
SAG using 5" media with several pebble ports (4") discharging smaller media for magnetic separation and media captured is being reloaded into the BM.
BM has 3" media.
Now the hard part. The media retains its spherical shape throughout deployment in both SAG and BM - repeating that all media remains round throughout deployment.
Small spherical ball populations grow in the BM promoting a better grind, SAG balls are not retained below say 3.5" leaving more space for new ore.
The question:
If the above was fully achieved, what would be the "estimated" improvement in efficiencies? I have no evidence (data) to quantify the media saving or the additional room for more ore. I also do not have an estimate of savings due to media remaining round and not losing shape. I do have one (and only one) reference for this - unused media additions to BM were reduced by 25% as this is the estimated media additions won from used round SAG media.
Our aim was always to reduce costs and increase efficiencies in existing SAG/BM circuits. The HPGR and VM alternative does read well - and calculating this new efficiencies in comparison with the SAG/BM best operating scenario is probably relevant?
Any thoughts or questions would be most welcome.

Sugar Watkins
1 year ago
Sugar Watkins 1 year ago

The technical and financial comparison «BM vs. VM», based on grinding test results, is not a problem. The comparison «SAG vs. HPRG» is more complex and requires the good knowledge of the entire grinding circuit and its phases, including the material classification and its handling. The experimental checking of the compatibility with the ore with HPRG process is obligatory. As an example, please look at the Juan remark related to the clay content. In 2011 - 2012, I developed a similar preliminary economic assessment (PEA) for our project Roche Bay Magnetite Deposit, Nunavut Canada. Based on the RB ore characteristics, production capacity and study and test results, I selected the final scenario: Primary Crushing (from 1000 mm to 175 mm); Wet FAG/SAG (from 175 mm to 1.6 mm) followed by VM grinding (from 1.6 mm to 30 µm).

The selection of the FAG/SAG as a primary grinding phase has been the result of the study of the efficiency of entire flowsheet, including magnetic separation and sulfide flotation. In spite of the higher profitability of the HPRG in comparison with FAG/SAG, the major financial impact of the efficiency of the wet ore classification and magnetic separation, on the global process efficiency, has been the final criterion of the selection of the FAG/SAG variant.

Carl Jenkins
1 year ago
Carl Jenkins 1 year ago

How much (t/d) you adding balls 5' to the SAG

At beginning all balls 100% 5'. After 20...25 days summary size is follow: 20% from 125mm to 100mm, 50% 100-80mm, 10% 80-40mm, 20% -40mm

Balls -50mm will be have cubic shape. itnecessary unload. Magnetic separator don’t apply (the unloading balls very coarse). May be apply gravity jig.

After you unload -50mm balls summary size in sag will be: 125...100mm 50%, 100...50mm 50%. All balls in sag will be have spherical shape. Unloading balls in ball mill after one day will be spherical too.

I chose variant a. but the number of VM are many. Do you tell me the actual data of arctic mill?

Dizzy Flores
1 year ago
Dizzy Flores 1 year ago

Usual SAG consumption milling mid hardness ores (say WI 16) can expect 300-400gr per tonne of ore milled. Harder ores and media breakage a take these numbers well over the KG per tonne consumption. We have seen (first hand) in access of 3kg/mt, due to unnecessary media breakage due to incorrect media grade/size selection.

Collectively all of us have the task to look closely in comparing new designs in equipment, meanwhile we also have the task to "optimize" existing equipment.
Strict SAG operation does not include grinding only crack the ore and discharge with only short transitional time in the primary mill. However, now we are all seeing higher ball charges and ore levels in these primary mills. Targeting a smaller size particle (P80) is the reasoning most likely for this? Taking some work from the secondary mill(s) for an improvement in the final grind follows that same logic.
Round media in the SAG mill in such cases then makes sense as round media in traditional SAG operation is no great gain (as no grinding is planned).
Your percentages for media size seasoned charge distribution is similar to our own, but with pebble porting those smaller sizes should have been mostly discharged.
SAG discharge being roughly classified by a screen to BM and oversize then magnetic separation and then to cyclone classification with overflow to the BM and underflow back to the SAG.
Keep an acceptance of media retaining its shape throughout deployment and read further.
BM operation using size/grade media blend to target the final grind - if the BM is at high efficiency then either throughput or grind can be adjusted with media selection - but not both. Recent site visits that we have had enabled both throughput and grind to be enhanced as efficiencies were low.
Sorry to write these long comments about SAG/BM circuits but the point is that many circuits can be improved greatly and so may be a more accurate circuit to compare with the HPGR/VM alternative. The best thing that I see (please correct me if I am wrong), with HPGR/VM scenario is that there is far less chance of getting such a circuit operating incorrectly - simplicity is its greatest gift?

Carmen Ibanz
1 year ago
Carmen Ibanz 1 year ago

An HPGR has less things to look at, compared to a SAG or Ball mill, when trying to optimize its performance. The residence time is really short in this machine and it is really easy to take a sample to see what the outcome of a change was.

In contrast, when you change the grind media distribution in a SAG or Ball Mill you need to wait a much larger period of time until you see the final result. By that time the ore may be different, liners may be at different state of their useful life, you may have new operators, etc.

Dizzy Flores
1 year ago
Dizzy Flores 1 year ago

After 70 plus comments, to original question,

We should be ready to agree on some items that we, as a group, can accept as valid.
1. HPGR - VM is more energy efficient PROVIDED that the ore body remains very consistent.
2. HPGR - VM requires a lower skill set for operation.
3. HPGR - VM will trend out instantly if adjustments (rebuild) are required.

SAG - BM circuits do come in second for the above. However HPGR, does not have the ability to deal with ore type, feed size, hour to hour changes.

It seems to be EFFICIENCY Versus FLEXIBILITY being the SIMPLE answer to the question?

Carl Jenkins
1 year ago
Carl Jenkins 1 year ago

Well I chose for my big project the HPGR.

Ace Levy
1 year ago
Ace Levy 1 year ago

I remarked you have selected the following grinding concept:

1. HPRG (from 50 mm to 3 mm size), as a primary grinding step;
2. VM grinding (from 3 mm to 150 µm; in one stage), as a final grinding phase.

Your choice seems to be right at this phase of the project (low OPEX and even lower CAPEX than SAG and 2 BM stages), but the tests are required, as follows:

• Laboratory and pilot HPRG tests, the last in closed circuit, aiming to qualify the response of your ore to the HPRG process; The tests will be developed on 1.2…1.5 t crushed ore, 50 mm size; I suggest you to select the P80 3 mm size instead 2 mm (HPRG process) in order to increase the dry screening efficiency, consequently, in order to reduce the screening CAPEX and OPEX and HPGR throughput;
• VM test on 60...70 kg ore, F80 3mm, necessary to the accurate quantification of the VM number/units; It is preferable, in order to reproduce the commercial operating conditions, to submit to VM test a sample resulted from the HPRG pilot test.

Concerning the VM number, my estimation, based on your information (BM WI = 12.6 kWh/t; F80 3 mm; P80 150 µm, is the following:

Production Capacity 3600 t/h

VM3000, 9 units, VM4500, 7 units or VM6000 (available from January 2015), 5 units.

Production Capacity 4000 t/h

VM3000, 10 units, VM4500, 7 units or VM6000 (available from January 2015), 5 units.

If you need the rough CAPEX and power consumptions required by the 6 options mentioned above, please contact me. For the current phase of your project (probably Preliminary Economic assessment or Scoping Study), the accuracy degree of my estimate is higher than the accuracy degree of your study. The VM tests will give you the very accurate VM number and power consumption. For the development of the tests, you can contact the METSO (VM) and Thyssen Krupp Polysius regional representative, or contact me in order to introduce you to METSO and TKP concerned people.

Marshal Dienes
1 year ago
Marshal Dienes 1 year ago

In conclusion:

Selection -- Type of grinding system----
i. For very hard ores SAG mill. For soft ores and all FLOTATION TECHNOLOGY use HPGR.
ii. Before taking decision consult Experts in Mineral processing for FLOTATION.
iii. Do compare results of metallurgy for both.
ivDo cost benefit analysis.
v. Develop operator knowledge in controlling OPERATING PARAMETERS as per design, not as per fancy ideas given by many. Redesign if you fail to get results.( from the company who has supplied equipment.)
vi. Many times ore characteristics changes as you go deep in the mine, or change location.

Carmen Ibanz
1 year ago
Carmen Ibanz 1 year ago

For very hard ores you may not want to use a SAG circuit. With hard I mean "competent" ores. there is a correlation between hard and competent ores if we talk about macro fractures.

If moisture in the ore is high and if the ore has too much clays , then HPGR may not be the right machine for the task.

Sugar Watkins
1 year ago
Sugar Watkins 1 year ago

FYI, the HPGRs accept max. 8% moisture. The clay is a major problem. It is possible to use HPGRs, but a prealable ore washing and, after, its dewatering are required.

Zander Barcalow
1 year ago
Zander Barcalow 1 year ago

If moisture in the ore is high and if the ore has too much clays, then HPGR may not be the right machine for the task

For all problems we have solutions. Clays in ore is 1st washed to -2mm, and +2mm. +2mm in INDIA will dry so fast that it will have hardly 6% moisture. You are giving examples of site specific. Yes you are correct if this ore is in cold countries at <10 C or where it is difficult to dry naturally. It will add to cost for drying +2 mm.
2.Let us not get confused with this subject. For some it is good for some it is not---This is based on situation, temperature, nature of ore, Flotation problems, Leaching problems, etc.
3.Anyone who want to use please do tests, and calculate cost benefit analysis form all points of view 100% then decide. It is only to know and gain more knowledge and each ones experience. Good to learn all have given reasonably good suggestions. All suggestions are very valuable and thought provoking.
4. We had very good time to learn many new concepts, practical solutions.
We need not stop at one point and use age old technologies. New technologies are equally good. Let us learn how best we can make use of them.
5.Our priority is Mineral conservation, recovery, yield, simple operation, cost after installation.
6.Today Metals have become so costly that CAPEX and OPEX can be absorbed in profit.

Dizzy Flores
1 year ago
Dizzy Flores 1 year ago

TIGHTEST control of ore particle size throughout the "pit to product" flow sheet is agreed by all of us, as paramount. After looking at this forum contributor list, and their credentials, we can take this as 100% incontrovertible fact.

We have also discussed how that this can be done best - right on topic.
Contributors have generously shared detailed firsthand information and also references to clarify points made. This has been invaluable to us all, it has certainly improved my own knowledge for one.

As mentioned earlier, we have also broadly categorized the benefits and shortcomings of both SAG and HPGR selection. HPGR for "efficiency" and SAG for "flexibility".

Agreed that this is a generalization, BUT we have, I think, given project designers a great "starting point" for initial assessment of equipment selection. The exercise has been totally worth the effort and we should all be pleased of the exchange. I do hope a balanced paper can be drafted on this very question, as it does have universal benefit.

Suggest we all keep these new friendly contacts on hand, as we should now have some new "go to guys" in specific disciplines.
I am glad to have exchanged some wisdom with all of you.

Marshal Meru
1 year ago
Marshal Meru 1 year ago

Adaptation of HPGR as fourth stage size reducer for high clay and bit over critical moisture content feed ore is difficult because of the increase in force of cohesion for adhered fine particles on roll. This may spoil the plant reliability and operator's confidence. The same seems to be addressed by pre-screening for fines reduction, pre washing to reduce sticky adhered fine clays then dry and etc. HPGR consumes less energy for the same duty condition because

a. It works on breaking the particle to particle cleavages. The surface coatings took place at the time of ore body evolution in metamorphism may not damage to the greater extent as it happens in case of SAG - Ball mill combination. This activity may be desired or less desired for some onward separation process based on surface coatings. HPGR generates narrow band PSD particles.
b. The SAG- Ball mill combination works on impacting and attrition action and the particles residence time increases with its size and density combination. This will generate slimes of high density valuable mineral particle, which is undesired activity and consumes unrequired energy. The hydro cyclones in efficiency will also add up and increase in close circuits grinding operation.

Size reduction is a liberation process which needs to be designed or focussed to felicitate more on wards separation process results. Off course selected combination essentially needs to strengthen plant reliability and economy.

David
1 year ago
David 1 year ago

Paul Morrow
1 year ago
Paul Morrow 1 year ago

The HPGR has earned a place in front end processing by being the best cost option for particle resizing, if the ore body has fixed and reliable characteristics - even I accept this and we are a ball supplier! I am content however, that the SAG mill still has greater utility in its flexibility to handle a wide range, or changing feed ore characteristic.

The 4 lines above, for my part, say all that needs to be said.

In the early releases there was too much expectation in HPGR ability to handle some great variations in an particular ore bodies. Careful planning of what goes in HPGR gives you the fixed and reliably fixed discharge result. I think Boddington (Australia) is happy with current results over the other option of 3 X 40' SAG mills, this will be one to research for sure.

SAG does not need careful planning should circumstances prevent that luxury, any size, any type of ore can be "dealt" with.

Bill Fraser
1 year ago
Bill Fraser 1 year ago

New technologies generally take years to get traction with mining companies but once they do get something good they learn to use it well. It always needs a senior operations person to champion the cause before there is "traction". Then that is just step #1....

We have had some new developed products accepted by large mining houses and not by the smaller ones and vice versa....so it just depends on the company people to embrace new things or to even get them to consider....I have found the copper guys to be more interested in exploring things lately...(probably something to do with USD3/Pd?).

Unterstarm
1 year ago
Unterstarm 1 year ago

Read:

http://911Metallurgist.com/C/HPGR-SAG/Compare%20HPGR%20vs%20SAG%20USING%20THE%20SMC%20TEST-TO%20PREDICT%20COMMINUTION%20CIRCUIT%20PERFORMANCE.pdf

http://911Metallurgist.com/C/HPGR-SAG/USE%20OF%20THE%20Mia%2C%20Mib%2C%20Mih%2C%20Mic%20IN%20PREDICTING%20COMMINUTION%20CIRCUIT%20SPECIFIC%20ENERGY%20of%20HPGR%20vs%20SAG%20.pdf

http://911Metallurgist.com/C/HPGR-SAG/Predicting%20SAG-AG%20Mill%20and%20HPGR%20Specific%20Energy.pdf

http://911Metallurgist.com/C/HPGR-SAG/HPGR%20vs%20SAG%20about_smc_test.pdf

Please join and login to participate and leave a comment.

BUY Laboratory & Small Plant Process Equipment

We have all the laboratory and plant equipment you need to test or build/operate your plant.

ENTER our Mining Equipment' Store

We Sell EQUIPMENT for all types of Mineral Treatment PROCESSES and Laboratory Testing needs

Have a Mineral Processing QUESTION?

Come in, ask your question

911Metallurgist Community Forums

Talk to other metallurgists and be helped.

Need ENGINEERING Services or Plant TROUBLESHOOTING?

We can IMPROVE ALL PLANTS / Mineral Processing Engineering & LABORATORY Ore Testing

911Metallurgy Engineering

Contact us for process engineering, metallurgical investigations, plant optimization, plant troubleshooting, needs. WE “FIX” METALLURGY.